Ranveer

Full Stack Engineer Intern , Data science , machine learning

Talent

Kolkata600 INR/hourMember since 19 May 2026

Data sciencemachine learningBackend DevelopmentAi EngineerDeep Learning

Hire this person

Send a job offer directly to this candidate

Data sciencemachine learningBackend DevelopmentAi EngineerDeep Learning

About

Building backend services using FASTAPI, designing APIs and managing MongoDB for AI-driven workflows, while keeping data flow smooth between services and the frontend. Working on integrating AI models into the application, handling data pipelines and ensuring responses are fast and reliable in real-time use cases. Deploying and managing services using Docker and AWS, setting up containers and environments so the system is stable and easier to scale , also worked on building ai agent by langchain , langgraph framework , also know about fine tuning of llm by qlora , lora , peft technique , know about some deep learning concept like ANN ,RNN , LSTM , GRU , Transformer : also created some rag pipeline for chatbot at internship and for my personal project Gen-ai product recommendation system , another ml project customer churn prediction model on a telecom company data --> in which i used to different boosting model like XGBOOST , LIGHTBGM , CATBOOST , ANN and evaluate the model by formula 0.65* RECALL + 0.35* F1 SCORE

Experience

Full Stack Engineer Intern at Point9 AI (2026-04 – Present)

Building backend services using FASTAPI, designing APIs and managing MongoDB for AI-driven workflows, while keeping data flow smooth between services and the frontend.
Working on integrating AI models into the application, handling data pipelines and ensuring responses are fast and reliable in real-time use cases.
Deploying and managing services using Docker and AWS, setting up containers and environments so the system is stable and easier to scale.

Software Development Intern at CountryBean (2025-04 – 2025-07)

Developed scalable AI chatbots leveraging RAG (FAISS), prompt engineering, and LangGraph-based agent workflows, enabling context-aware and tool-augmented interactions.
Architected persistent memory systems using SQLite + summarization pipelines, improving response relevance and reducing latency overhead in multi-turn conversations.
Implemented function-calling agents with dynamic tool usage (session management, user data handling, retrieval orchestration) and fine-tuned LLaMA (QLoRA) for domain-specific response optimization.